R Notebook

# load data.table for faster operations
library(data.table)
library(tidyverse)

# use fread for fast reading of data csv files
  # ColumnNameCleaner.rmd should be run first
cases <- fread("Data/time_series_covid19_confirmed_US.csv")
vaccineCounty <- fread("Data/Vaccination/COVID-19_Vaccinations_in_the_United_States_County.csv")
#vaccineJurisdiction <- fread("Data/Vaccination/COVID-19_Vaccinations_in_the_United_States_Jurisdiction.csv")
pop <- fread("Data/Population/co-est2019-alldata.csv")
distancing <- fread("Data/Distancing/Unacast_Social_Distancing_Grades.csv")
masking <- fread("Data/Masking/mask-use-by-county.csv")

# preview the data tables
head(cases)

head(vaccineCounty)

#head(vaccineJurisdiction)
head(distancing)

head(masking)

head(pop)

# either create or fix the FIPS column using available data
cases[, FIPS := sprintf("%05d", cases[,FIPS])]
pop[, FIPS := sprintf("%02d%03d", pop[,STATE], pop[,COUNTY])]
distancing[, FIPS := sprintf("%05d", county_fips)]
masking[, FIPS := sprintf("%05d", COUNTYFP)]

# drops unneeded columns
cases <- cases[, !c("V1", "UID", "iso2", "iso3", "code3", "CombinedKey")]
# selects the most recent population estimate (2019) and drops what will be redundant columns
years2019 <- grep("^([^0-9]*)$|2019", colnames(pop))
pop <- pop[, ..years2019][, !c("SUMLEV", "STATE", "COUNTY", "STNAME", "CTYNAME")]
# drops what will be redundant columns
distancing <- distancing[, !c("OBJECTID", "state_fips", "state_name", "county_fips", "county_name")]
masking <- masking[, !c("COUNTYFP")]

# set the FIPS column as the key for faster data.table operations
# time series
setkey(cases, FIPS)
setkey(vaccineCounty, FIPS)
setkey(distancing, FIPS)
# not time series
setkey(masking, FIPS)
setkey(pop, FIPS)

pivot_wider(vaccineCounty, names_from = Date, values_from = )

csv <- cases[pop,][distancing,][masking,]
write.csv(cases[pop,][distancing,][masking,], "Data/casesPopDistancingMasking.csv", row.names = FALSE)
csv